[SOLVED] Problem when reading a .med file created by Saturne

youdestock · 2018-03-12 03:56:02

Hey there,
I'm trying to reading a .med file created by Saturne, I want to read the "TempC (Boundary)" from the "Fluid_result.med" file, and project it on the boundary of the "solid_mesh.med" file for further calculation, here's my code :

MAFLUIDE = LIRE_MAILLAGE(FORMAT='MED', NOM_MED='Fluid domain', UNITE=20)
MASOLIDE = LIRE_MAILLAGE(FORMAT='MED', UNITE=2)
MOFLUIDE = AFFE_MODELE(
AFFE=_F(MODELISATION=('3D', ), PHENOMENE='THERMIQUE', TOUT='OUI'),
INFO=1,
MAILLAGE=MAFLUIDE
)
MOSOLIDE = AFFE_MODELE(
AFFE=_F(MODELISATION=('3D', ), PHENOMENE='THERMIQUE', TOUT='OUI'),
MAILLAGE=MASOLIDE
)
mou_mat = DEFI_MATERIAU(
ELAS=_F(ALPHA=1.2e-05, E=2e+11, NU=0.3, RHO=7870.0),
THER=_F(LAMBDA=60.0, RHO_CP=400)
)
MATER = AFFE_MATERIAU(
AFFE=_F(GROUP_MA=('vol_tube', 'vol_mould'), MATER=(mou_mat, )),
MAILLAGE=MASOLIDE
)
L_INST = DEFI_LIST_REEL(DEBUT=1.0, INTERVALLE=_F(JUSQU_A=100.0, NOMBRE=100))
lir_RE = LIRE_RESU(
FORMAT='MED',
FORMAT_MED=_F(NOM_CHAM='PRES', NOM_CHAM_MED='TempC (Boundary)'),
INFO=2,
MODELE=MOFLUIDE,
PROL_ZERO='OUI',
TOUT_ORDRE='OUI',
TYPE_RESU='EVOL_CHAR',
UNITE=81
)
RES_PROJ = PROJ_CHAMP(
METHODE='COLLOCATION',
MODELE_1=MOFLUIDE,
MODELE_2=MOSOLIDE,
RESULTAT=lir_RE,
TOUT_ORDRE='OUI',
VIS_A_VIS=_F(GROUP_MA_1=('face_group_1', ), GROUP_MA_2=('face_group_1', ))
)
load = AFFE_CHAR_THER(EVOL_CHAR=RES_PROJ, MODELE=MOSOLIDE)
resther = THER_NON_LINE(
CHAM_MATER=MATER,
ETAT_INIT=_F(INST=0.0, VALE=21.0),
EXCIT=_F(CHARGE=load),
INCREMENT=_F(LIST_INST=L_INST),
MODELE=MOSOLIDE
)
FIN()

Unfortunately, it doesn't work. Always a problem when reading the result from the MED file.
Could anyone help me ?
The Fluid_result.med and the solid_mesh.med will be attached blow.

youdestock · 2018-03-12 04:06:56

Med File - part1

Attachments:
Mesh.part1.rar, Size: 1.8 MiB, Downloads: 595

youdestock · 2018-03-12 04:08:41

Med File - part2

Attachments:
Mesh.part2.rar, Size: 1.8 MiB, Downloads: 500

youdestock · 2018-03-12 04:10:43

Med File - part3

Attachments:
Mesh.part3.rar, Size: 1.76 MiB, Downloads: 546

Thomas DE SOZA · 2018-03-13 13:42:54

Hello,

The solid mesh MED file is OK.

The fluid results MED file has a problem or at least contains unrecognised outputs for the MED format.

Attempt at converting the file to MED 3.0.8 version

[NS=dsp0780510:2810]$ ConvertMEDFileTo30.py Fluid_results.med 
Traceback (most recent call last):
  File "/opt/salome-meca/INTEGR/V8_INTEGR/tools/Medcoupling-V8_4_0/bin/ConvertMEDFileTo30.py", line 47, in <module>
    ConvertTo30(nameOfMEDFile)
  File "/opt/salome-meca/INTEGR/V8_INTEGR/tools/Medcoupling-V8_4_0/bin/ConvertMEDFileTo30.py", line 33, in ConvertTo30
    mfd=ml.MEDFileData(nameOfMEDFile)
  File "/opt/salome-meca/INTEGR/V8_INTEGR/tools/Medcoupling-V8_4_0/lib/python2.7/site-packages/MEDLoader.py", line 28949, in __init__
    this = _MEDLoader.new_MEDFileData(*args)
MEDLoader.InterpKernelException: constructor MEDFileFields(fileName) : file 'Fluid_results.med' at pos #2 field has name 'boundary zone id' but the type of field is not in [MED_FLOAT64, MED_INT32, MED_FLOAT32] !

Attempt at opening the file in the MED module of Salome-Meca

>>> import pvsimple as pvs
>>> 
>>> import medcalc
>>> medcalc.medconsole.setConsoleGlobals(globals())
>>> import MEDCALC
>>> 
>>> from medcalc.medconsole import saveWorkspace, cleanWorkspace
>>> from medcalc.medconsole import putInWorkspace, removeFromWorkspace
>>> from medcalc.medconsole import accessField
>>> from medcalc.medconsole import getEnvironment, ls, la
>>> 
>>> source_id = medcalc.LoadDataSource('/home/E50954/Fluid_results.med')
Traceback (most recent call last):
  File "<input>", line 1, in <module>
  File "/opt/salome-meca/INTEGR/appli_V8_INTEGR/lib/python2.7/site-packages/salome/medcalc/medio.py", line 24, in LoadDataSource
    handler = dataManager.loadDatasource(filename)
  File "/opt/salome-meca/INTEGR/appli_V8_INTEGR/lib/python2.7/site-packages/salome/MEDDataManager_idl.py", line 212, in loadDatasource
    return _omnipy.invoke(self, "loadDatasource", _0_MEDCALC.MEDDataManager._d_loadDatasource, args)
omniORB.CORBA.UNKNOWN: CORBA.UNKNOWN(omniORB.UNKNOWN_UserException, CORBA.COMPLETED_MAYBE)
>>> source_id
Traceback (most recent call last):
  File "<input>", line 1, in <module>
NameError: name 'source_id' is not defined
>>> presentation_id = medcalc.MakeMeshView(medcalc.GetFirstMeshFromDataSource(source_id), viewMode=MEDCALC.VIEW_MODE_REPLACE)
Traceback (most recent call last):
  File "<input>", line 1, in <module>
NameError: name 'source_id' is not defined
>>> presentation_id
Traceback (most recent call last):
  File "<input>", line 1, in <module>
NameError: name 'presentation_id' is not defined
>>>

You should review the output and maybe remove the field name "boundary zone id".

TdS

youdestock · 2018-03-13 16:25:05

Thanks a lot for your reply Thomas,

I checked my fluid mesh, and I generate a new one which has been attached below.(time step=40, output result every 5 steps)
I check the new Fluid mesh file with Salome-meca, it seems correct.
But, Aster-study still cannot read it.
Is there's something wrong with my Code? if so, will you show me the correct LIRE_RESU code and maybe PROJ_CHAMP?

P.S. I have no idea of how to delete "boundary zone id", I think it is automatically created by code_Saturne.

Cordialement,

Attachments:
re_Fluid_mesh.med, Size: 914.13 KiB, Downloads: 652

Thomas DE SOZA · 2018-03-14 13:48:35

Hello,

The new fluid mesh is now correct in the sense that it can be read by the MED module of Salome-Meca (this was not the case before).

Regarding the error you get when launching the calculation, there are several things you need to fix:

the fluid mesh file on which you will define a "fluid model" in code_aster should be the boundary not the fluid domain, as it is the only part to be in contact with the solid domain
the LIRE_RESU part is incomplete as the component of the 'TempC (boundary)' field ("scalar") is not what is expected by code_aster ('TEMP'). You should use NOM_CMP and NOM_CMP_MED in LIRE_RESU/FORMAT_MED
importing an external field through LIRE_RESU with the EVOL_CHAR type is limited to a few examples detailed in the manual of LIRE_RESU [U7.02.01] : for temperature, you can only import T_EXT (with component TEMP) and COEF_H (with component H) which enable the use of an exchange boundary condition in thermics

You should also review both of the following testcases : ZZZZ137B for LIRE_RESU imports of temperature fields and ZZZZ215C for LIRE_RESU imports of pressure fields from Code_Saturne.

Good luck,

TdS

youdestock · 2018-03-15 18:56:18

Hello Thomas,
Thank you for your advices.
I made several changes for my code:

1. the "bonudary" part is imported:

MAFLUIDE = LIRE_MAILLAGE(
  FORMAT='MED', 
  NOM_MED='Boundary', 
  UNITE=2
)

2. change the NOM_CHAM equals to 'T_EXT' in LIRE_RESU. But still very confuse after reading the manual LIRE_RESU [U7.02.01].
In your second advice, you suggest me to use NOM_CMP and NOM_CMP_MED in LIRE_RESU/FORMAT_MED and you also said that the component should be 'TEMP' for aster_study. so,

How to display the component in the med file? if it is not TEMP, How am I suppose to change it?
Not very clear about the definition of Opérandes NOM_CMP/NOM_CMP_MED/NOM_CHAM_MED (NOM_CHAM is clear, it equals to 'T_EXT' when reading temperature). Page 9 du manuel LIRE_RESU [U7.02.01], "T_EXT: Carte de température extérieure, composante TEMP", si j'ai bien compris, c'est-a-dire,
NOM_CHAM='T_EXT',
NOM_CHAM_MED='TempC (Boundary)' ____c'est le nom du champ qu'on veut lire dans le fichier .med
NOM_CMP=('TEMP', )____c'est le composante 'TEMP'
mais c'est quoi le 'NOM_CMP_MED', pouvez-vous m'expliquer le 'NOM_CMP_MED'?

Si j'ecrit le code comme ça:

lir_RE = LIRE_RESU(
  FORMAT='MED', 
  FORMAT_MED=_F(
    NOM_CHAM='T_EXT', 
    NOM_CHAM_MED='TempC (Boundary)', 
    NOM_CMP=('TEMP', ), 
    NOM_CMP_MED=('TempC (Boundary)', )
  ), 
  MODELE=MOFLUIDE, 
  PROL_ZERO='OUI', 
  TOUT_ORDRE='OUI', 
  TYPE_RESU='EVOL_CHAR', 
  UNITE=3
)

Il y a une erreur

JDC.py : ERREUR A L'EXECUTION - INTERRUPTION
>> JDC.py : DEBUT RAPPORT
CR d'execution de JDC en MIXTE
   !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
   ! <S> Exception utilisateur levee mais pas interceptee. !
   ! Les bases sont fermees.                               !
   ! Type de l'exception : error                           !
   !                                                       !
   !   -> Composante inconnue pour la grandeur.            !
   !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
fin CR d'execution de JDC en MIXTE

>> JDC.py : FIN RAPPORT
EXECUTION_CODE_ASTER_EXIT_12073=1

 <I>_EXIT_CODE = 1

si je supprime NOM_CMP et NOM_CMP_MED (comme l'example ZZZZ137B), erreur suivant va afficher:

!----------------------------------------------------------------!
   ! <A> <POSTRELE_56>                                              !
   !                                                                !
   !  la composante Scalar n'est pas une composante de TEMP_R       !
   !                                                                !
   !                                                                !
   ! Ceci est une alarme. Si vous ne comprenez pas le sens de cette !
   ! alarme, vous pouvez obtenir des résultats inattendus !         !
   !----------------------------------------------------------------!

Last edited by youdestock (2018-03-15 18:58:23)

Thomas DE SOZA · 2018-03-15 22:56:42

Hello,

The result file in MED you retrieved from Code_Saturne has the following characteristics:

it contains a field name "TempC (Boundary)" ==> so NOM_CHAM_MED='TempC (Boundary)'
this field contains only one component named "Scalar" ==> so NOM_CMP_MED='Scalar'

You want to use that field in code_aster to input a temperature load:

the imposed field name for the EVOL_CHAR type of load is "T_EXT" ==> so NOM_CHAM='T_EXT'
the imposed component name for the T_EXT field in code_aster est "TEMP" ==> so NOM_CMP='TEMP'

TdS

youdestock · 2018-03-22 04:08:13

Sorry for the delay, I'm quite busy these days. thank you for the reply.
with your useful and professional instruction I manage to read the .med file.
Best reagrds,

Last edited by youdestock (2018-03-22 18:14:16)

Attachments:
re_Fluid_mesh.med, Size: 914.13 KiB, Downloads: 588

jacob · 2022-09-19 14:35:05

Hello,
how to read and project temperature from whole domain?
I am trying to use T_EXT, but it does not work.

jacob · 2022-09-20 10:45:15

if I do

loadtran = LIRE_RESU(identifier='7:1',

FORMAT='MED',

FORMAT_MED=_F(NOM_CHAM='TEMP',

NOM_CHAM_MED='resther_TEMP'),

MODELE=CFDmodel,

PROL_ZERO='OUI',

TOUT_ORDRE='OUI',

TYPE_RESU='EVOL_THER',

UNITE=20)

on a model computed by code_aster, it works.
But on results from code_saturne it doesn't.

DEBUT(LANG='EN')

Mmesh = LIRE_MAILLAGE(UNITE=3)

CFDmesh = LIRE_MAILLAGE(FORMAT='MED',
UNITE=81)

Mmodel = AFFE_MODELE(AFFE=_F(MODELISATION='3D',
PHENOMENE='MECANIQUE',
TOUT='OUI'),
MAILLAGE=Mmesh)

CFDmodel = AFFE_MODELE(AFFE=_F(MODELISATION='3D',
PHENOMENE='THERMIQUE',
TOUT='OUI'),
MAILLAGE=CFDmesh)

mater = DEFI_MATERIAU(ELAS=_F(E=200000000000.0,
NU=0.3))

fieldmat = AFFE_MATERIAU(AFFE=_F(MATER=(mater, ),
TOUT='OUI'),
MAILLAGE=Mmesh)

loadtran = LIRE_RESU(FORMAT='MED',
FORMAT_MED=_F(NOM_CHAM='TEMP',
NOM_CHAM_MED='TempC'),
MODELE=CFDmodel,
PROL_ZERO='OUI',
TOUT_ORDRE='OUI',
TYPE_RESU='EVOL_THER',
UNITE=81)

unnamed0 = PROJ_CHAMP(MODELE_1=CFDmodel,
MODELE_2=Mmodel,
PROJECTION='OUI',
PROL_ZERO='OUI',
RESULTAT=loadtran)

IMPR_RESU(RESU=_F(RESULTAT=unnamed0),
UNITE=2)

FIN()

║ <A> <POSTRELE_56> ║
║ ║
║ the component Scalar is not a component of TEMP_R ║
║ ║
║ ║
║ This is a warning. If you do not understand the meaning of this ║
║ warning, you can obtain ║
║ unexpected results!

EDIT:
If anyone finds it useful.

DEBUT(LANG='EN')

Mmesh = LIRE_MAILLAGE(UNITE=20)

CFDmesh = LIRE_MAILLAGE(FORMAT='MED',

UNITE=4)

Mmodel = AFFE_MODELE(AFFE=_F(MODELISATION='3D',

PHENOMENE='MECANIQUE',

TOUT='OUI'),

MAILLAGE=Mmesh)

CFDmodel = AFFE_MODELE(AFFE=_F(MODELISATION='3D',

PHENOMENE='THERMIQUE',

TOUT='OUI'),

MAILLAGE=CFDmesh)

mater = DEFI_MATERIAU(ELAS=_F(E=200000000000.0,

NU=0.3))

fieldmat = AFFE_MATERIAU(AFFE=_F(MATER=(mater, ),

TOUT='OUI'),

MAILLAGE=Mmesh)

timestep = DEFI_LIST_REEL(DEBUT=0.0,

INTERVALLE=_F(JUSQU_A=100.0,

NOMBRE=50))

loadtran = LIRE_RESU(FORMAT='MED',

FORMAT_MED=_F(NOM_CHAM='TEMP',

NOM_CHAM_MED='TempC',

NOM_CMP=('TEMP', ),

NOM_CMP_MED=('Scalar', )),

LIST_INST=timestep,

MODELE=CFDmodel,

PROL_ZERO='OUI',

TYPE_RESU='EVOL_THER',

UNITE=4)

unnamed0 = PROJ_CHAMP(MODELE_1=CFDmodel,

MODELE_2=Mmodel,

PROJECTION='OUI',

PROL_ZERO='OUI',

RESULTAT=loadtran)

IMPR_RESU(RESU=_F(RESULTAT=unnamed0),

UNITE=2)

FIN()

But I had to apply on my results Cell Data to Point Data in Paraview, and use those results.

Last edited by jacob (2022-09-20 13:02:12)

#1 2018-03-12 03:56:02

[SOLVED] Problem when reading a .med file created by Saturne

#2 2018-03-12 04:06:56

Re: [SOLVED] Problem when reading a .med file created by Saturne

#3 2018-03-12 04:08:41

Re: [SOLVED] Problem when reading a .med file created by Saturne

#4 2018-03-12 04:10:43

Re: [SOLVED] Problem when reading a .med file created by Saturne

#5 2018-03-13 13:42:54

Re: [SOLVED] Problem when reading a .med file created by Saturne

#6 2018-03-13 16:25:05

Re: [SOLVED] Problem when reading a .med file created by Saturne

#7 2018-03-14 13:48:35

Re: [SOLVED] Problem when reading a .med file created by Saturne

#8 2018-03-15 18:56:18

Re: [SOLVED] Problem when reading a .med file created by Saturne

#9 2018-03-15 22:56:42

Re: [SOLVED] Problem when reading a .med file created by Saturne

#10 2018-03-22 04:08:13

Re: [SOLVED] Problem when reading a .med file created by Saturne

#11 2022-09-19 14:35:05

Re: [SOLVED] Problem when reading a .med file created by Saturne

#12 2022-09-20 10:45:15

Re: [SOLVED] Problem when reading a .med file created by Saturne

Board footer